Re: [docbook-apps] adding additional data to the olink database

Hi Tom,

The olink database is just an XML file 
generated by XSLT templates.  It certainly could be adapted to capture 
additional information.  You would have to figure out where in the olink 
data element structure you would put the new data (elements, attributes), and 
then customize some of the templates in common/targets.xsl to generate those 
elements.  I don't think the new data would conflict with the existing 
olink processing, since it calls specific element and attribute 
names.

 

Bob Stayton
Sagehill Enterprises
DocBook 
Consulting


 

 

  
----- Original Message ----- 

  
From: 
   
  

  
To: docbook-apps 

  
Sent: Thursday, October 13, 2005 6:53 
  AM

  
Subject: [docbook-apps] adding additional 
  data to the olink database

  

I am currently using the olink 
  database file for some additional uses in my website (I use it to build a 
  sitemap, and to list the contents of particular directories). I want to use it 
  to generate a listing of all articles in a particular directory - which I can 
  do fine. The problem is that I need to access a little more information about 
  each document than what is provided in the olink sitemap. I would like to have 
  access to the abstract and author information. Currently I am using XSL's 
  document() function to open up each article to extract the required data but I 
  feel this is a little slow an inefficient, and I would rather avoid having to 
  open up each individual article just to access this little bit of extra 
  information. 

Some of my articles 
  use XIncludes to include dynamic content, so whenever I do 
  "document('path/to/docbook/document.xml')/article/articleinfo/abstract" 
  xsltproc processes all the Xincludes in the entire document, which puts a 
  heavy load on my server and slows things down a lot (because the server is 
  having to generate the dynamic content for each article, just to load it into 
  memory). 

Is there a way to make 
  the olink database generation stylesheets include extra information? 
  

or 

Is there a way to tell xsltproc to not process XIncludes (although I am 
  not using xsltproc directly, but through AxKit - a Cocoon style dynamic xml 
  transformation pipeline framework, so I do not think I will be able to tinker 
  with xsltproc very easily)? 

many thanks

-- 
Tom Kirkpatrick
Web Developer - Virus 
  Bulletin